Approximating electronically excited states with equation-of-motion linear coupled-cluster theory

TL;DR

This paper introduces new perturbative methods based on linear coupled-cluster theory for approximating excited states, showing they closely match more accurate methods and experimental data for small organic molecules.

Contribution

It presents novel EOM-LCCD and EOM-LCCSD methods using perturbation theory, offering computationally efficient alternatives to canonical EOM-CCSD for excited states.

Findings

Excellent agreement with EOM-CCSD in state ordering and energies

Acceptable quantitative accuracy for excitation energies

Validated on 25 small organic molecules

Abstract

A new perturbative approach to canonical equation-of-motion coupled-cluster theory is presented using coupled-cluster perturbation theory. A second-order M{\o}ller-Plesset partitioning of the Hamiltonian is used to obtain the well known equation-of-motion many-body perturbation theory (EOM-MBPT(2)) equations and two new equation-of-motion methods based on the linear coupled-cluster doubles (EOM-LCCD) and linear coupled-cluster singles and doubles (EOM-LCCSD) wavefunctions. This is achieved by performing a short-circuiting procedure on the MBPT(2) similarity transformed Hamiltonian. These new methods are benchmarked against very accurate theoretical and experimental spectra from 25 small organic molecules. It is found that the proposed methods have excellent agreement with canonical EOM-CCSD state for state orderings and relative excited state energies as well as acceptable quantitative agreement for absolute excitation energies compared with the best estimate theory and experimental spectra.